Seasonal rotation features of wind vectors and application to evaluate monsoon simulations in AMIP models

Abstract

A new concept, the directed angle, is intro- duced to study seasonal rotation regimes of global wind vectors and annual variability of monsoon. Compared with previous studies on using angles between wind vectors, this concept better describes the daily variations of both rotation direction and rotation amplitude of wind vector. According to the concept, six categories of wind vector rotation with seasonal cycle in the global have been detected and classified as follows: (1) Clockwise to counter-clockwise (CTCC) rotation; (2) Counter-clock- wise to clockwise (CCTC) rotation; (3) Full clockwise (FC) rotation; (4) Full counter-clockwise (FCC) rotation; (5) Stable style; (6) Unstable style. Generally, wind vec- tors in monsoon regions rotate in forms of the first four styles. Moreover, the rotation direction and rotation amplitude of wind vectors have regional differences, and different monsoon subsystems possess different rotation styles for wind vectors in an annual cycle. For instance, the South Asian monsoon follows the CCTC rotation, while the East Asian monsoon follows the FCC rotation. The CTCC rotation is seen in the South China Sea. Both the West Africa and the South Indo-China Peninsula are covered by the FC rotation. Therefore, the directed angle is able to describe the evolution of wind vectors on a daily scale, which provides a new clue for spatio-temporal information about wind vector variation and model eva- luation. Using the new concept, this study aims at evaluating the model outputs of eight AGCMs of AMIP in the Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report (AR4). Compared to the cor- responding observations, most models are able to simulate the global rotation regimes of wind vectors reasonably well, however very little skill is shown in the monsoon rotation styles of some models, especially in the South China Sea and West Africa. Moreover, the simulations differ mostly from observations during the transitional season.